The electric power system comprises generation, transmission, and distribution assets, which deliver power to loads. With the introduction of renewable resources, the electric power system faces a number of constraints which favor the addition of storage assets.
The principal constraint on an interconnected grid is the need to maintain the frequency and voltage by balancing variations in generation and demand (load). Failure to maintain the voltage or frequency within specifications causes protective relays to trip in order to protect generators, transmission and distribution assets from damage. Because of the interconnected dynamic electrical grid, underfrequency or undervoltage trips can cause a cascade of other trips, potentially leading to widespread blackouts.
Traditionally, electric utilities or the independent system operators managing electrical grids maintain a power generation reserve margin that can respond to changes in load or the loss of a generating unit or transmission line serving the load. These reserves are managed and scheduled using various planning methods, including day-ahead forecasts, dispatch queues that may be ordered based on generation cost, and generation ramp-rates, transmission constraints, outages, etc. The spinning generation units, that is, those that are operating, then respond to generation load control signals.
Many renewable resources are intermittent in nature, including wind farms, central station solar thermal or solar photovoltaic (PV) plants, and distributed photovoltaic systems including those on rooftops. These can produce power only when the resource is available, during daylight for solar, and when the wind is blowing for wind, leading to seasonal and diurnal production variations, as well as short-term fluctuations due to calms, gusts, and clouds. Gusts that exceed wind turbine ratings may cause them to trip with a sudden loss of full generation capacity. Deployment of these renewable systems as both central and distributed generating resources results in fluctuations in both the generation of power to be transmitted and the demand for power, since the distributed PV offsets load.
Base load is usually provided by large central station nuclear, hydroelectric or thermal power plants, including coal-fired steam plants (Rankine cycle) or gas-fired Combined Cycle Combustion Turbine plants (open Brayton air cycle with closed Rankine steam bottoming cycle). Base-load units often have operating constraints on their ramp rates (Megawatts per minute) and Turn-Down (minimum Megawatts), and startup from cold steel to rated load requires several hours to several days depending on the type and size of generating asset. Accordingly, a different class of load following power plants is also deployed in the electric power system, to complement the base load units. Generally, these load following units are less efficient in converting thermal energy to electrical energy.
This conversion efficiency is often expressed as a Heat Rate with units of thermal energy needed to produce a kilowatt-hour (kw-hr) of electricity [British Thermal Unit (BTU) per kw-hr in the U.S., kiloJoules (kJ) per kw-hr elsewhere]. The thermal equivalent of work is 3413 BTU/kw-hr or 3600 kJ/kw-hr, which represents 100% efficiency. Modern combined cycle power plants at full load rated conditions may have heat rates as low, for example, as 6000 kJ/kw-hr. Modern gas turbine peaking plants (e.g., General Electric LM6000-PC SPRINT) can achieve a full load rated condition heat rate of 9667 kJ/kw-hr (HHV). It is important to note that gas turbine heat rates increase rapidly away from rating conditions, and at part load in hot conditions the actual heat rate may be twice the rated Heat Rate.
It is of course desired to deliver electricity to customers at the lowest possible cost. This cost includes the amortization and profit on invested capital, the operating and maintenance (O&M) expense, and the cost of fuel. The capital amortization (and return on capital, in the case of regulated generators) is applied to the capacity factor (fraction of rated generation) to arrive at the price ($ per Megawatt-hour) associated with the fixed capital expense. The Heat Rate multiplied by the fuel cost determines the contribution of the variable fuel consumption to the electricity price. The O&M expenses generally have some combination of fixed and variable expenses, but are insignificant compared to capital and fuel for central station plants. Generating units have different mixes of fixed and variable expenses, but presumably were believed to be economic at the time they were ordered.
In order to deliver low cost electricity to a customer, it is necessary to operate the capital intensive units at high capacity, subject to fuel cost, in order to spread the capital cost across many kw-hr. Contrariwise, it is necessary to minimize the operation of units with high marginal operating cost (due to high Heat Rate, Fuel Cost or O&M). This was indeed the planning assumption for procurement of the existing fleet of generators.
The Renewable resources gather ‘free’ fuel, so their cost of generation is dominated by the amortization of the capital needed to gather and convert this energy into electricity. In order to profitably build and operate a Renewable power plant, it should have as high a capacity factor as may be practically realized. Similarly, the fuel-efficient base load generation should operate at high capacity factor, both to amortize the capital expense, and because its operating characteristics induce higher fuel or O&M costs (per unit of generation) when operated intermittently or at part load.
The increasing penetration of renewable generation with variable generation characteristics is challenging the traditional dispatch order and cost structure of the electric generation system. In practice, utility scale solar power plants without storage are limited to Capacity Factors of about 25%, and wind farms seldom exceed 50%. This capacity may not coincide with demand, and may be suddenly unavailable if the sun or wind resource is reduced by local weather. For example, if wind resources are available at periods of low demand, base load units must either ramp down or shut-down or the wind resource must be curtailed. If the wind is not curtailed, then less efficient peaking units may be needed to provide ramp flexibility that the large base-load units cannot provide in case of gusts or calms. Likewise the widespread deployment of solar power generation is depressing the need for generation during daylight hours, but large ramp rates as the sun rises and sets can currently only be met by gas fired peaking plants. Ironically, this will result in displacement of low-cost, high efficiency base-load units in favor of high cost, low-efficiency peaking units, with a concomitant increase in greenhouse gas releases.
For environmental, energy security, cost certainty and other reasons, renewable energy sources are preferred over conventional sources. Demand Response techniques, which attempt to reduce the instantaneous load demand to achieve balance between generation and consumption, are analogous to a peaking generation unit. Another approach is deployment of (e.g., large scale) energy storage systems to mediate the mismatch between generation and consumption.
Storage systems are alternately charged to store energy (e.g., using electric power), and discharged to return the energy as power to the electric grid. The technical characteristics of energy storage systems include:                the Capacity, or quantity of energy that can be stored and returned, measured in MW-hours;        the Round Trip Efficiency (RTE), or fraction of the energy delivered to the storage system that is returned to the grid;        the Power rating, or rate in MW at which the system can be charged or discharged (Power is often symmetric, though this is not necessary, or even desirable);        the Lifetime, which is typically the number of Charge/Discharge cycles.        
Pumped Storage Hydroelectricity (PSH) employs a reversible pump-turbine with two water reservoirs at different elevations. When excess energy is available, it is used to pump water from a lower to an upper reservoir, converting the electricity into potential energy. When electricity is needed, the water flows back to the lower reservoir through a hydro-turbine-generator to convert the potential energy into electricity. Pumped hydro storage is suitable for grid scale storage and has been used for many decades in electrical grids around the world. PSH has a Round Trip Efficiency (RTE) of 70% to 80% and can be deployed at Gigawatt scale with many days of potential storage. In addition to high RTE, PSH does not generate greenhouse gases during operation. Deployment of PSH requires availability of suitable locations for the construction of dams and reservoirs, and evaporative water loss may be an issue in some locations.
Compressed Air Energy Storage (CAES) stores pressurized air that is subsequently expanded in an engine. Commercially deployed CAES stores the air in large underground caverns such as naturally occurring or solution-mined salt domes, where the weight of overburden is sufficient to contain the high pressures. The RTE for CAES may be relatively low. The 110 MW McIntosh CAES plant in the US state of Alabama reportedly has a RTE of only 27%, for example. Several near-isothermal CAES technologies are also under development with reported RTE of 50% or greater, using pressure vessels for storage.
Many energy storage technologies are being developed and deployed for end-use loads or distribution level capacities, at power levels from a few kilowatts to several megawatts. These approaches typically employ batteries with a variety of chemistries and physical arrangements.
There is a need for high efficiency energy storage that is not dependent on geological formations, and which can be deployed at scales of tens to hundreds of megawatts to complement the existing generation and transmission assets.